Luminmycins A-C, cryptic natural products from Photorhabdus luminescens identified by heterologous expression in Escherichia coli

Article abstract of DOI:10.1021/np300444e

The 18 kb silent luminmycin biosynthetic pathway from Photorhabdus luminescens was cloned into a vector by using the newly established linear-linear homologous recombination and successfully expressed in Escherichia coli. Luminmycins A-C (1-3) were isolated from the heterologous host, and their structures were elucidated using 2D NMR spectroscopy and HRESIMS. Luminmycin A is a deoxy derivative of the previously reported glidobactin A, while luminmycins B and C most likely represent its acyclic biosynthetic intermediates. Compound 1 showed cytotoxicity against the human colon carcinoma HCT-116 cell line with an IC₅₀ value of 91.8 nM, while acyclic 2 was inactive at concentrations as high as 100 μg/mL.

Full text of DOI:10.1021/np300444e

Note
pubs.acs.org/jnp
Luminmycins A−C, Cryptic Natural Products from Photorhabdus
luminescens Identified by Heterologous Expression in Escherichia coli
Xiaoying Bian,† Alberto Plaza,† Youming Zhang,*,‡ and Rolf Muller*,†
̈
†Helmholtz Institute for Pharmaceutical Research Saarland, Helmholtz Centre for Infection Research and Department of
Pharmaceutical Biotechnology, Saarland University, P.O. Box 15115, 66041 Saarbrucken, Germany
̈
‡Gene Bridges GmbH, Building C2 3, Saarland University, 66123 Saarbrucken, Germany
̈
S
* Supporting Information
ABSTRACT: The 18 kb “silent” luminmycin biosynthetic pathway from Photorhabdus luminescens was cloned into a vector by
using the newly established linear−linear homologous recombination and successfully expressed in Escherichia coli. Luminmycins
A−C (1−3) were isolated from the heterologous host, and their structures were elucidated using 2D NMR spectroscopy and
HRESIMS. Luminmycin A is a deoxy derivative of the previously reported glidobactin A, while luminmycins B and C most likely
represent its acyclic biosynthetic intermediates. Compound 1 showed cytotoxicity against the human colon carcinoma HCT-116
cell line with an IC50 value of 91.8 nM, while acyclic 2 was inactive at concentrations as high as 100 μg/mL.
he increasing number of bacterial genome sequences have
Trevealed an enormous amount of cryptic or silent
secondary metabolite biosynthetic pathways, outnumbering by
far the known natural products from these species.1
Heterologous expression of biosynthetic pathways in a well-
characterized host is a rational approach to mine this hidden
resource of natural products, especially for slow-growing
sources. However, the time-consuming and labor-intense
cloning and engineering of large biosynthetic gene clusters
restricts the application of this approach.2 Recently, we
reported the full-length RecE and its partner RecT in E. coli-
mediated linear plus linear homologous recombination, which
can circumvent the limitations of cloning large gene clusters
without construction and screening of genomic libraries.3 By
using this technique, 10 unknown secondary metabolite
biosynthetic gene clusters were directly cloned from the
genomic DNA of Photorhabdus luminescens subsp. laumondii
TT01 (DSM15139)4 into plasmids under inducible promoters.
Two of these gene clusters, plu3263 and plu1881-plu1877, were
successfully expressed in E. coli, and their assigned products
were designated as luminmides A/B and luminmycin A (1),
respectively.3 It is worth noting that compound 1 was not
previously detected in HPLC-MS chromatograms of extracts
obtained from P. luminescens, demonstrating that the cryptic
biosynthetic gene cluster plu1881−plu1877 was activated by
heterologous expression.3 The structure of 1 was proposed on
the basis of comparative biosynthetic analysis and HRMS
fragmentation with glidobactin A.3,5,6 Furthermore comparative
HPLC/MS analysis between tetracycline-induced and unin-
duced extracts of E. coli strain Nissle 19177 containing
plu1881−plu1877 indicated the presence of more luminmycin
derivatives. However, structure elucidation and purification of
these derivatives were hampered by low-yield production in the
heterologous host. To improve the yield, tetracycline-inducible
promoter PtetO was changed to a constitutive promoter by
replacement of tetracycline-inducible repressor gene tetR with
chloramphenicol-resistant gene cmR via Red/ET recombina-
tion.8,9
Luminmycins A−C (1−3) were absorbed onto Amberlite
XAD 16 adsorber resin that had been added to a 12 L batch
fermentation of E. coli Nissle harboring the luminmycin gene
cluster. The products were eluted from the resin with MeOH
and further purified by Sephadex LH-20 chromatography and
repeated semipreparative RP-HPLC, yielding compounds 1
(3.0 mg), 2 (0.5 mg), and 3 (0.4 mg).
Received: June 25, 2012
Published: August 21, 2012
© 2012 American Chemical Society and
American Society of Pharmacognosy
1652
dx.doi.org/10.1021/np300444e | J. Nat. Prod. 2012, 75, 1652−1655

Journal of Natural Products
Note
Table 1. NMR Spectroscopic Data for Luminmycin B (2)
a
b
c
d
position
δC, type
δH (J in Hz)
7.63, m
HMBC
11, 12
ROESY
11, 12
1-NH
2
3
173.1, C
37.9, CH2 2.14, m
2.27, m
2, 4, 5
4, 5
4
68.7, CH
48.0, CH
3.84, m
3, 7, 13
3, 13
3, 13
5
3.81, m
3, 4, 7,13
5, 7, 13
6-NH
7.55, d (8.5)
3, 5, 8, 13
7
8
9
170.8, C
51.9, CH
4.26, m
7, 9, 10, 1′ 6, 9, 10
7, 8, 10, 11 1, 8, 10, 12,14
31.1, CH2 1.52, m
1.70, m
10
11
21.9, CH2 1.31, m
26.3, CH2 1.51, m
8, 9, 11, 12 8, 9, 12,
10, 12
1, 8, 10, 12,
14
12
38.4, CH2 2.74, m
15.7, CH3 1.00, d (6.8)
7.91, d (8.2)
10, 11
4, 5
1, 10, 11
Me-13
14-NH
1′
3, 4, 5, 6
8, 9, 1′
8, 9, 10, 2′, 4′
169.8, C
2′
58.2, CH
4.32, dd (4.6, 7.9) 1′, 4′, 5′, 1″ 14, 5′
3′-NH
4′
5′
1″
2″
7.95, d (8.1)
2′, 4′, 5′, 1″ 4′, 5′, 2″
66.5, CH
3.98, m
1′, 5′
2′, 4′
14, 2′, 5′
2′, 3′, 4′
The HRESIMS of 1 showed an ion peak at m/z 505.3378 [M
+ H]+ corresponding to a molecular formula of C27H44N4O5.
Detailed analysis of the 2D NMR data of 1 (see Table S1) in
comparison to that of the previous reported glidobactin A5,6
clearly indicated 1 to be a 10-deoxy derivative of glidobactin A.
This result corroborates the proposed structure previously
reported.3 The HRESIMS spectra of 2 and 3 showed major ion
peaks at m/z 541.3604 [M + H]+ and m/z 497.3345 [M + H]+
corresponding to molecular formulas of C27H48N4O7 and
C25H45N4O6, respectively. These values indicated that the
molecular weight of 2 was 36 amu higher than 1, while 3 was 44
amu lower than 2. The 2D NMR data for 2 closely resembled
those of 1, differing only by the signals corresponding to the
macrocycle ring. In particular 2 showed an additional
oxymethine group (δC‑4 68.7, δH‑4 3.84) and a proton signal
at δ 7.63 (2H, m) ascribable to a terminal NH2 group (Table
1). It was also evident that the olefinic carbons at δ 6.23 and
6.78 in 1 were absent in 2. HMBC correlations from the methyl
protons at δ 1.00 to the carbon resonances at C-4 and from the
methylene protons at δ 2.14 and 2.27 to the carbonyl at δ 173.1
established the presence of a 4-amino-3-hydroxypentanoic acid
(Ahp) residue (Table 1, Figure 1). Long-range correlations
from the methine proton at δ 3.81 and the amide proton at δ
7.55 to the lysine carbonyl at δ 170.8 linked the Ahp residue to
the lysine (Figure 1), thus establishing the structure of 2 as a
seco-hydrated derivative of 1. The 2D NMR data for 3 closely
resembled those of 2 with the exception that resonances
belonging to the Ahp were replaced by resonances belonging to
an alanine residue (Table S2). The geometries of all of the
double bonds in the dodecadienoic acid (Dda) residue of 2 and
19.3, CH3 1.05, d (6.3)
165.5, C
122.8, CH
139.5, CH
6.16, m
1″, 3″, 4″
3′, 3″
4″, 5″
3″
7.00, dd (11.4,
15.1)
1″, 2″, 4″,
5″
4″
5″
128.4, CH
142.0, CH
6.18, m
2″, 3″, 5″,
6″
3″, 6″, 7″
3″, 6″, 7″
6.10, m
3″, 4″, 6″,
7″
6″
7″
8″−9″
10″
11″
12″
32.1, CH2 2.13, m
28.2, CH2 1.38, m
28.2, CH2 1.25, m
31.1, CH2 1.24, m
21.9, CH2 1.28, m
13.6, CH3 0.86, t (6.8)
3″, 4″, 7″
5″, 6″
5″, 7″
5″, 6″
11″, 12″
10″, 12″
10″, 11″
10″, 11″
10″, 12″
11″
a
Recorded at 175 MHz; referenced to residual DMSO-d6 at δ 39.51
b
ppm. Recorded at 700 MHz; referenced to residual DMSO-d6 at δ
c
2.50 ppm. Proton showing HMBC correlation to indicated carbon.
Proton showing ROESY correlation to indicated proton.
d
Figure 1. Key correlations of HMBC, ROESY, and TOCSY of
luminmycin B (2). Lys: lysine; Thr: threonine; Ahp: 4-amino-3-
hydroxypentanoic acid; Dda: 2(E),4(E)-dodecadienoic acid.
3
3 were established as E based on the large JH−H values (∼15
Hz). Additional evidence confirming the structure of 2 was
obtained from ESIMS/MS experiments. Fragmentation of the
major ion peak at m/z 541 [M + H]+ displays an intense ion at
m/z 523 [M + H − H2O]+ and several fragments of small
intensity at m/z 505 [M + H − 2H2O]+, 390 [M + H − H2O −
Ahp]+, 346 [M + H − H2O − Dda]+, 280 [M + H − Lys −
Ahp]+, 262 [M + H − Thr − Dda]+, and 244 [M + H − H2O −
Thr − Dda]+. Therefore the MS2 fragmentation patterns were
in total agreement with the structure proposed for 2 by NMR.
The absolute configurations of threonine (Thr), lysine (Lys),
and alanine (Ala) were determined as L by MS-detected
chromatographic comparison of the derivatives of the acid
hydrolysate (6 N HCl, 90 °C, 16 h) of 1 and 3 with L-FDLA (1-
fluoro-2,4-dinitrophenyl-5-L-leucinamide)10,11 and D-FDLA de-
rivatives of amino acid standards (Table S3, Figures S1 and S2).
Because these three compounds are biosynthesized by one gene
cluster, the absolute configurations of the amino acids in 2
1653
dx.doi.org/10.1021/np300444e | J. Nat. Prod. 2012, 75, 1652−1655

Journal of Natural Products
Note
mode at a scan range of m/z = 100−1400, auto MSn. Chromatographic
conditions: Luna RP C18 column, 100 × 2 mm, 2.5 μm particle size,
and precolumn C18, 8 × 3 mm, 5 μm; solvent gradient (with solvents
A [H2O and 0.1% formic acid] and B [CH3CN and 0.1% formic acid])
from 5% B at 2 min to 95% B within 20 min, followed by 3 min with
95% B at a flow rate of 0.4 mL/min.
should be the same as 1 and 3. Low amounts of 2 prevent the
assignment of the configuration at C-4 by NMR or chemical
derivatizations. Inspection of Plu1880 indicates that the
hydroxy group at C-4 is formed by reduction of a keto group
by a ketoreductase (KR) domain. Nevertheless prediction of
the configuration at C-4 was hampered by the lack of an LDD
motif and a highly conserved typtophan in this KR domain
(Figure S3).12 Compounds 2 and 3 are assumed to be
intermediates of the luminmycin assembly line hydrolyzed from
the carrier proteins. The possible biosynthetic pathway of 1−3
was proposed according to the known luminmycin biosynthesis
(Figure S4).3 Compound 3 is hydrolyzed from the second
thiolation (T) domain of Plu1880 directly without the
involvement of the last PKS module, leading to the absence
of a PKS extender unit (malonyl-CoA). The last PKS module
contains functional KR and dehydratase (DH) domains;
however it is possible to generate β-hydroxyacyl or β-ketoacyl
intermediates by skipping the DH or KR domains.13
Compound 2 is the hydrolysis product from β-hydroxyacyl-S-
T lacking dehydration by DH domain in the last PKS module
(Figure S4). Thus, the identification of 2 and 3 supports the
previously hypothetical biosynthetic pathway of luminmycin A
in vivo.
Moreover we evaluated the cytotoxic and antifungal activity
of 1 and 2. Luminmycin A (1) is a potent inhibitor of tumor
cell proliferation in vitro, inhibiting the growth of the tumor cell
line HCT-116 with an IC50 value of 91.8 nM, which is weaker
than that of glidobactin A (IC50 33.3 nM). The acyclic
derivative luminmycin B (2) was found inactive at concen-
trations as high as100 μg/mL. Glidobactin A shows antifungal
activity against Candida albicans with an MIC value of 2 μg/
mL, while 1 and 2 show no activity. The loss of cytotoxic and
antifungal activity in the open-ring luminmycin B fits with the
mechanism of proteasome inhibition, because the α,β-
unsaturated carbonyl moiety in the 12-membered-ring system
binds covalently to the hydroxy group of the active site amino
N-terminal threonine of the proteasome.14
Extraction and Isolation. E. coli Nissle strain 19177 containing
pGB-PtetO-plu1881-plu1877-cm was cultivated in 12 L of M9
medium15 supplemented with 10 μg/mL chloramphenicol and 2%
XAD 16 resin (after 2 days' incubation) at 30 °C for 5 days. The resin
was collected by sieving, washed with H2O twice, and then extracted
stepwise with MeOH (5 L). The extract was concentrated under
reduced pressure, followed by suspension in MeOH and extraction
with n-hexane to degrease. The resulting MeOH extract (0.8 g) was
fractionated initially on a Sephadex LH-20 column (100 × 2.5 cm)
using MeOH as a mobile phase, and 55 fractions were obtained.
Fractions containing compounds 1−3 were subjected to a semi-
preparative reversed-phase HPLC system (Jupiter Proteo C12, 250 ×
10 mm, 4 μm, DAD at 254 nm) with isocratic 75% MeOH/H2O with
0.05% TFA to yield 1 (3.0 mg, tR = 21.5 min). Compounds 2 (0.5 mg,
tR = 23.9 min) and 3 (0.4 mg, tR = 27.3 min) eluted with isocratic 60%
MeOH.
Luminmycin A (1): white, amorphous solid; [α]20 −8.1 (c 0.36;
D
MeOH); UV (MeOH) λmax (log ε) 261 nm (4.85); IR (film) νmax
3314, 2925, 1644, 1522, 1213, 999 cm−1; for NMR data see Table S1;
HRESIMS m/z 505.3378 [M + H]+ (calcd for C27H45N4O5,
505.3384).
Luminmycin B (2): colorless, amorphous solid; [α]20D −3.4 (c 0.06;
MeOH); UV (MeOH) λmax (log ε) 261 nm (4.76); IR (film) νmax
3294, 2933, 1671, 1540, 1209, 1140, 1000, 726 cm−1; for NMR data
see Table 1; HRESIMS m/z 541.3604 [M + H]+ (calcd for
C27H49N4O7, 541.3596).
Luminmycin C (3): colorless, amorphous solid; [α]20D −1.1 (c 0.02;
MeOH); for NMR data see Table S2; HRESIMS m/z 497.3345 [M +
H]+ (calcd for C25H45N4O6, 497.3334).
Determination of Amino Acid Configurations. Approximately
0.2 mg of 1 and 3 were hydrolyzed with 6 N HCl (0.8 mL) at 90 °C
for 16 h. These solutions were evaporated to dryness, and the residue
was dissolved in 100 μL of H2O. To each half-portion (50 μL) were
added 1 N NaHCO3 (20 μL) and 1% 1-fluoro-2,4-dinitrophenyl-5-L-
leucinamide (L-FDLA or D-FDLA in acetone, 100 μL), and the mixture
was vortexed and incubated at 37 °C for 60 min. The reaction was
quenched by the addition of 2 N HCl (20 μL) and evaporated to
dryness. The residues were resuspended in 300 μL of CH3CN, and
about 10 μL of each solution of FDLA derivatives was analyzed by
HPLC/MS. The HPLC/MS conditions are the same as that in the
general experimental procedure. The retention times of the Marfey-
derivatized amino acids are summarized in Table S3 and Figures S1
and S2. Retention times of the FDLA-derivatized authentic standards
are L-Thr 13.4 min, D-Thr 14.7 min, L-allo-Thr 13.5 min, D-allo-Thr
14.4 min, m/z 414.2 [M + H]+; L-Lys (mono-α) 11.5 min, D-Lys
(mono-α) 11.2 min, m/z 441.3 [M + H]+; L-Lys (di) 18.7 min, D-Lys
(di), 19.2 min, m/z 735.4 [M + H]+; L-Ala 14.6 min, D-Ala 15.5 min,
m/z 384.2 [M + H]+.
In conclusion, we have isolated two new peptides,
luminmycins B and C, by heterologous expression of the
genes plu1881−plu1877 in E. coli. The previously identified
luminmycin A was also isolated, and its structure corroborated
by NMR. Additionally, luminmycin A shows interesting
cytotoxic activity toward a human colon tumor cell line.
Thus, expression of a cryptic biosynthetic gene cluster in a
heterologous host represents a powerful approach to mine the
enormous biosynthetic capabilities of microorganisms.
EXPERIMENTAL SECTION
■
General Experimental Procedures. Optical rotations were
measured on a Perkin-Elmer polarimeter. UV data were recorded on
a NanoDrop 2000C spectrophotometer (Thermo Scientific) using
MeOH as solvent. IR data were recorded on a Perkin-Elmer Spectrum
100 FT-IR spectrometer. NMR spectra were recorded on a Bruker
Ascend 700 MHz spectrometer equipped with a CryoProbe system.
The samples were measured in DMSO-d6, and the solvent peaks were
used as references (δC 39.51, δH 2.50 ppm). Data acquisition,
processing, and spectral analysis were performed with standard Bruker
software, TopSpin 3.0. Chemical shifts are given in ppm and coupling
constants in Hz. HRESIMS data were recorded on an LTQ-Orbitrap
(Thermo Scientific), and molecular formulas were identified by
including the isotopic pattern in the calculation (Xcalibur). Analytical
RP-HPLC was carried out on an Agilent 1100 HPLC system equipped
with a UV diode-array detector and a Bruker Daltonics HCTultra ESI-
MS ion trap instrument operating in positive and negative ionization
ASSOCIATED CONTENT
■
S
* Supporting Information
Supplementary experimental section, NMR spectroscopic data
and advanced Marfey’s results for 1 and 3, biosynthetic
pathways and copies of 1D and 2D NMR spectra of 1−3. This
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*(Y.Z.) Tel: +49 681 30270236. Fax: +49 681 30270202. E-
mail: youming.zhang@genebridges.com. (R.M.) Tel: +49 681
1654
dx.doi.org/10.1021/np300444e | J. Nat. Prod. 2012, 75, 1652−1655

Journal of Natural Products
Note
30270201. Fax: +49 681 30270202. E-mail: rom@mx.uni-
saarland.de.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
X.B. is indebted to China Scholarship Council (CSC) for a
Ph.D. scholarship. We thank J. Herrmann and T. Hoffmann
from R.M.’s laboratory for bioactivity tests and high-resolution
MS measurements. Research in the laboratory of R.M. was
funded by the Deutsche Forschungsgemeinschaft (DFG) and
the Bundesministerium fur Bildung und Forschung (BMBF).
̈
Y.Z. in Gene Bridges was partially funded by the BMBF
(MiPro).
REFERENCES
■
(1) Bode, H. B.; Muller, R. Angew. Chem., Int. Ed. 2005, 44, 6828−
6846.
̈
(2) Wenzel, S. C.; Muller, R. Curr. Opin. Biotechnol. 2005, 16, 594−
̈
606.
(3) Fu, J.; Bian, X.; Hu, S.; Wang, H.; Huang, F.; Seibert, P. M.; Plaza,
A.; Xia, L.; Muller, R.; Stewart, A. F.; Zhang, Y. Nat. Biotechnol. 2012,
30, 440−446.
(4) Duchaud, E.; Rusniok, C.; Frangeul, L.; Buchrieser, C.; Givaudan,
A.; Taourit, S.; Bocs, S.; Boursaux-Eude, C.; Chandler, M.; Charles, J.
F.; Dassa, E.; Derose, R.; Derzelle, S.; Freyssinet, G.; Gaudriault, S.;
Medigue, C.; Lanois, A.; Powell, K.; Siguier, P.; Vincent, R.; Wingate,
V.; Zouine, M.; Glaser, P.; Boemare, N.; Danchin, A.; Kunst, F. Nat.
Biotechnol. 2003, 21, 1307−1313.
(5) Oka, M.; Nishiyama, Y.; Ohta, S.; Kamei, H.; Konishi, M.; Miyaki,
T.; Oki, T.; Kawaguchi, H. J. Antibiot. 1988, 41, 1331−1337.
(6) Oka, M.; Yaginuma, K.; Numata, K.; Konishi, M.; Oki, T.;
Kawaguchi, H. J. Antibiot. 1988, 41, 1338−1350.
(7) Sonnenborn, U.; Schulze, J. Microb. Ecol. Health Dis. 2009, 21,
122−158.
(8) Zhang, Y.; Buchholz, F.; Muyrers, J. P.; Stewart, F. A. Nat. Genet.
1998, 20, 123−128.
(9) Zhang, Y.; Muyrers, J. P. P.; Testa, G.; Stewart, A. F. Nat.
Biotechnol. 2000, 18, 1314−1317.
(10) Fujii, K.; Ikai, Y.; Mayumi, T.; Oka, H.; Suzuki, M.; Harada, K.
Anal. Chem. 1997, 69, 3346−3352.
(11) Fujii, K.; Ikai, Y.; Oka, H.; Suzuki, M.; Harada, K. Anal. Chem.
1997, 69, 5146−5151.
(12) Keatinge-Clay, A. T. Chem. Biol. 2007, 14, 898−908.
(13) Fischbach, M. A.; Walsh, C. T. Chem. Rev. 2006, 106, 3468−
3496.
(14) Groll, M.; Schellenberg, B.; Bachmann, A. S.; Archer, C. R.;
Huber, R.; Powell, T. K.; Lindow, S.; Kaiser, M.; Dudler, R. Nature
2008, 452, 755−758.
(15) Sambrook, J.; Russell, D. W. Molecular Cloning: A Laboratory
Manual; Cold Spring Harbor Laboratory Press: Cold Spring Harbor,
NY, 2001.
1655
dx.doi.org/10.1021/np300444e | J. Nat. Prod. 2012, 75, 1652−1655